Pulse Radiolysis of Neopentane in the Gas Phase

Abstract

The pulse radiolysis of gaseous neopentane has been investigated in the absence and presence of electron scavengers (SF₆, CD₃I, CCl₄). Deuterium labeling experiments show that the stable product molecules can be accounted for by (a) radical combination reactions involving mainly CH₃ and H; (b) hydride ion transfer reactions involving C₂H₃⁺, C₂H₅⁺, and C₃H₅⁺; (c) neutralization reactions of C₄H₉⁺ and C₅H₁₁⁺; and (d) unimolecular dissociation of the parent ion (C₅H₁₂⁺) and of electronically excited neopentane. Neutralization of the t-C₄H₉⁺ ion, which is the major positive ion in the system occurs as follows: (a) t-C₄H₉⁺ + e → i-C₄H₈ + H and (b) t-C₄H₉⁺ + e → 2CH₃ + C₃H₆. It is shown that C₅H₁₁⁺ produced in hydride ion transfer reaction C_nH_m⁺ + neo-C₅H₁₂ → C_nH_m+1 + C₅H₁₁⁺ (where C_nH_m⁺ = C₂H₃⁺, C₂H₅⁺, and C₃H₅⁺) rearranges to the CH₃C⁺(CH₃)CH₂CH₃ structure prior to neutralization. A detailed accounting of all products produced in the unimolecular and bimolecular reactions led to the conclusion that the ratio of neutral electronically excited molecules to parent ions (N_ex/N₊) is 0.28.

1. Introduction

Although nearly 500 papers dealing with pulse radiolysis have appeared in the literature, pulse radiolysis studies of alkanes in the gas phase are nonexistent, with the exception of two cursory studies of methane [1].¹ The reason for this is that it has become customary to associate direct detection techniques with pulse radiolysis experiments, and these techniques can not be readily applied to hydrocarbon gases. Actually, in the case of hydrocarbons (with the exception of aromatics), the absorption spectra of most of the fragment ions and free radicals, which are known to be present, have never been observed [2].

The limitations of optical absorption techniques for these systems do not, however, rule out pulse radiolysis as a powerful tool in the elucidation of the chemical changes brought about by high energy irradiation of hydrocarbon gases. Indeed, the pulse irradiation of judiciously selected hydrocarbons, or hydrocarbon-containing mixtures, followed by chemical analysis of the end products, may yield a wealth of information which will contribute to our overall understanding of radiolytic phenomena.

The main distinguishing feature of high dose rate pulse radiolysis experiments is the short lifetime of the charged species. Neutralization will be entirely homogeneous and reactions of ions with minor impurities or accumulated products will be negligible. Furthermore, the absence of long ionic chain reactions simplfies the interpretation of the initial unimolecular and bimolecular reactions.

The present study which is the first in a series devoted to the pulse radiolysis of alkanes deals with neopentane. This molecule was chosen for the present exploratory study, because of the relatively limited number of reactive intermediates produced and because the low dose rate radiolysis has been investigated [3]. In the discussion which follows, the unimolecular and bimolecular reactions contributing to the formation of the products will be examined, by taking advantage of the same diagnostic isotopic labeling techniques which were successfully employed in earlier low dose rate investigations [3].

2. Experimental Procedure

2.1. Pulsed Electron Beam Sources

Two different electron pulse instruments, the Febetron 705 and 706,² were used in this study. These pulse sources are capacitor-discharge type instruments which use the field emission effect to obtain a high current of electrons from fine tungsten needles in a vacuum tube. The Febetron 705 has a half-maximum pulse duration of about 55 ns at maximum output (35 kV charging voltage, 4000 G focusing field). The beam contains approximately 2 × 10¹⁵ electrons per pulse with a mean energy of 1.4 MeV. At lower charging voltages the half maximum pulse duration is somewhat longer (about 75 ns at 16 kV charging voltage) and there are fewer electrons per pulse. The electron beam has a diameter of 3–4 cm just outside the thin titanium window of the instrument. For a given diode tube, the pulse-to-pulse reproducibility is excellent with a maximum deviation from the average of about 6 percent and an average deviation of less than 3 percent [4].

The Febetron 706 has half maximum pulse duration of about 3 ns at maximum output (30 kV charging voltage). The beam contains approximately 2 × 10¹⁴ electrons per pulse with a nominal energy of 600 keV and it has a diameter of about 1.4 cm just outside the window of the tube. For a given diode tube the pulse-to-pulse fluctuations are less than 5 percent mean square deviation [5].

A few flash photolysis experiments were carried out in the course of this study using the flash apparatus constructed by W. Braun [6].

2.2. Radiation Cells

The radiation cells used in this work were cylindrical borosilicate glass vessels 10 cm long and 3.7 cm internal diameter with a volume of 106 cm³. Aluminum windows which were 4 mil thick were attached to each end of the cell with a vacuum epoxy cement. These cells were evacuated and filled through a PTFE valve and the reactants were condensed in a small finger at the bottom of the cell with liquid nitrogen. The outside of the cell was covered with aluminum foil which was grounded when the cell was irradiated by the electron pulse. These cells were very carefully positioned as close as possible in front of the titanium window of the Febetron.

In some of the early experiments cylindrical stainless steel cells were used which were 10 cm long and 4.7 cm internal diameter with a volume of 175 cm³. The 4 mil aluminum windows were attached with dairy clamps and viton O-rings. Essentially similar results were obtained with either the stainless steel or the borosilicate glass cells.

2.3. Materials and Analysis

Most of the chemicals used in this investigation were of research grade quality and except for degassing there was no further purification. Two different sources of neopentane were used, but there was no substantial difference in the results obtained with different samples. Neopentane-d₁₂ was purified on a gas chromatograph and checked on a mass spectrometer. It contained 8 percent C₅D₁₁H. The CD₃I contained 9 percent CD₂HI. The nitric oxide was stored over silica gel.

After the irradiation by the electron pulse(s), an aliquot of the mixture was injected onto a squalane column of a gas chromatograph with a flame ionization detector for quantitative analysis of the products. In those experiments in which the isotopic distributions of some of the products were determined, the hydrogen and methane were distilled at either liquid nitrogen or solid nitrogen temperatures. Then the remainder of the mixture was injected onto a silica gel column and the various products were trapped at the exit of the column in borosilicate glass spiral traps maintained at liquid nitrogen temperature and were analyzed on a mass spectrometer.

2.4. Dosimetry

Nitrous oxide at 1 atm pressure was used as the dosimetric gas using G(N₂) ⁼ 12.4 as determined by Willis et al. [4,7]. The relative stopping power per electron of neopentane to nitrous oxide or air was calculated from the tables of Berger and Seltzer [8] as 1.05 for 600 keV electron (Febetron 706) and 1.03 for 1.4 MeV electrons (Febetron 705). The dosimetry experiments were performed in the same cells that were used to irradiate the neopentane mixtures so that there is no error due to a difference in geometry of the cells. One pulse emitted by the Febetron 706 (3 ns pulse) operated at a charging voltage of 30 kV corresponded to a dose of 3.6 × 10¹⁹ eV/g. The dose per pulse (~ 60 ns) from the Febetron 705 was 18.2 and 5.8 × 10¹⁹ eV/g at charging voltages of 30 and 16 kV respectively.

3. Results

The yields of the measured products are given as molecules, M, of product X produced per ion pair, N₊, (M(X)/N₊) [9]. A W-value of 23.2 eV for neopentane was chosen to convert number of molecules per eV to the number of molecules per ion pair [10].

Table 1 gives all of the lower hydrocarbon products whose ion pair yields have been determined. In addition the ion pair yields of 2-methyl-l-butene (0.066), 2-methyl-2-butene (0.025) and 2,2-dimethylbutane (0.050) were determined. Addition of SF₆ increased the yields of 2-methyl-l-butene, 2-methyl-2-butene and 2,2-dimethylbutane to 0.09, 0.03 and 0.068 respectively. Minor products such as n-butane (< 0.001) and propane (~ 0.01) are not listed in table 1.

Table 1.

Pulse radiolysis of neopentane product yields

Pulse1 duration (ns)	Dose/ pulse eV/ g × 10−19	No. of pulses	Additive mol %	M/N+ (Ion pair yield)2
				h2	CH4	C2H2	c2h4	C2H6	C3H6	CH3C≡CH	CH2 =C= CH2	i-C4H10	i-C4H8
3	3.6	1		n.d.	1.09	0.052	0.115	0.304	0.205	0.041	0.030	0.007	0.837
3	3.6	2		.35	1.07	.051	.112	.307	.215	.043	.038	.009	.849
3	3.6	2	CC14–2.0	n.d.	.375	.039	.097	.293	.100	.022	.013	.027	.902
3	3.6	3	CD3I-2.0	n.d.	.496	.030	.100	.811	.107	.026	.015	.069	.738
60	18.2	1		n.d.	1.11	.052	.117	.293	.218	.050	.042	.012	.842
60	5.8	1		n.d.	1.06	.041	.108	.305	.212	.044	.033	.009	.830
60	18.2	1	SFe-0.5	.33	.591	.039	.112	.375	.126	.033	.018	.029	.947
60	5.8	2	CD3I-1.0	n.d.	.542	.031	.100	.742	.104	.025	.013	.067	.769
60	5.8	2	CD3I-3.0	n.d.	.501	.032	.106	.878	.101	.023	.012	.072	.764
									Relative1	fields3
Flash			none	n.d.	.516	.009	.065	.645	.210	.032	.036	.030	LOO
photolysis

¹Full width half maximum.

²Pressure neopentane = 162 torr.

³Pressure neopentane = 5.0 torr.

n.d. = not determined.

Tables 2 and 3 present isotopic analyses of hydrocarbon products produced in the pulse radiolysis of equimolar neo-C₅H₁₂−neo-C₅D₁₂ mixtures. The hydrogen was also analyzed mass spectrometrically. In the pure neopentane mixture it consisted of H₂ (50.1%); HD (26.2%) and D₂ (23.7%). Addition of 1 mol percent SF₆ did not alter this distribution in any appreciable way, giving H₂ (47.1%); HD (29.5%) and D₂ (23.4%).

Table 2.

Pulse radiolysis of neo-C₅D₁₂ − neo-C₅H₁₂ (1:1) isotopic compositions of ethane and methane

Dose/pulse eV/g × 10−19	Additive mole %	Ethane distribution (%)					M/N+
		c2h6	C2H5D	C2H3D3	C2D5H	C2D6	(C2H5+ + C2D5+)
Pulse radiolysis:
18.2	None	26.5	5.5	37.6	5.0	23.7	0.062
18.2	SFe-1.0	26.5	4.4	41.3	4.0	22.1	.060
3.6	NO-3.0	27.5	18.0	13.0	16.5	25.0	.047
Low Dose Rate
Radiolysis (ref. 3):
	NO-5.0	29.5	20.5		21.2	28.8	.062
		Methane distribution (%)					M/N+ (Molecular methane)
		CH4	CH3D	CH2D2	CHD3	CD4
Pulse radiolysis:
18.2	none	30.8	16.0	2.6	27.5	23.1	~ 0.11
18.2	SF6-1.0	35.2	13.0	2.0	22.5	27.3	.16
3.6	NO-3.0	42.6	7.3		12.2	37.9	.16
Low Dose Rate
Radiolysis (ref. 3):
	NO-5.0	44.3	4.3		5.2	46.2	.18

Pressure of neopentane: pulse radiolysis 160 torr, low dose rate radiolysis (ref. 3) 40 torr.

Table 3.

Pulse radiolysis of neo-C₅D₁₂ − neo-C₅H₁₂ (1:1) isotopic compositions of ethylene and propylene

Dose/pulse eV/g × 10−19	Additive mole %	Ethylene distribution (%)				M/N+ (C2H3+ + C2D3+)
		c2h4	C2H3D	C2D3H	C2D4
Pulse radiolysis:
18.2	None	48.1	7.3	7.0	37.9	0.032
3.6	NO-3.0	49.5	6.8	5.9	37.8	.028
Low dose rate
radiolysis (ref. 3):
	NO-5.0	45.1	6.0	6.2	42.7	.028
		Propylene distribution (%)				M/N+(C3H5+ + C3D5+)
		C3H6	C3H5D	C3D5H	C3D6
Pulse radiolysis:
18.2	None	41.6	5.4	5.2	41.7	0.044
3.6	SFe-1.0	40.3	8.2	7.9	38.1	.041
Low dose rate
radiolysis (ref. 3):
	NO-5.0	37.1	11.1	11.2	40.0	.048

Pressure of neopentane: pulse radiolysis 160 torr, low dose rate radiolysis (ref. 3) 40 torr.

Table 4 contains the isotopic analyses and ion pair yields of the hydrogen, methane, and ethane produced in the pulse radiolysis of neo-C₅H₁₂ − CD₃I mixtures. The propylene and isobutene which were analyzed as well, contained less than 2 percent of the deuterium labeled species.

Table 4.

Pulse radiolysis of neo-C₅H₁₂−CD₃I mixtures

cd3i mol %	M/N+ (Ion Pair Yields)							CH4corr /CD3Ha	[CH3]/[CD3] CD3CH3/2C2D6	C2H6corr1/2 /C2D61/2
	H2	HD	cd3h	CH4	C2D6	CD3CH3	C2H6
1	.30	.006	.184	.358	.147	.330	.254	1.08	1.12	1.18
3.1			.176	.325	.225	.413	.240	0.94	0.92	0.92
6.1	.32	.013	.184	.312	.280	.472	.259	0.83	.84	.86

Dose per pulse: 5.8 × 10¹⁹ eV/g.

^a${\text{CH}}_{{\text{4}}_{\text{corr}}}$: measured yield of CH₄ minus contribution due to reactions 3, 5, 6, and 9.

^b${\text{C}}_2{\text{H}}_{{\text{6}}_{\text{corr}}}$: measured yield of C₂H₆ minus contribution due to reaction 19.

4. Discussion

The emphasis in this discussion will be mainly on the chemical consequences of the neutralization of the positive ions. In order to obtain information about this, one must examine all of the free radical and ion-molecule reactions which occur during and immediately after the pulse. Only when end products resulting from these reactions are accounted for, can one proceed to derive concrete information concerning the chemical consequences of the encounters between the unreactive positive ions and the negatively charged species.

Passage of an electron beam through neopentane will lead to the production of electronically excited neutral molecules and of excited parent ions [3]. These species are expected to dissociate within 10⁻¹² to 10⁻¹³ s [11, 12], and the following dissociative processes have been invoked to account for the products observed in the low dose rate experiments.

$${\text{C}}_5{\text{H}}_{12}^{*}\to i-{\text{C}}_4{\text{H}}_8+{\text{CH}}_3+\text{H}$$ (1)

$$\to {\text{C}}_3{\text{H}}_6+{\text{CH}}_3+{\text{CH}}_3$$ (2)

$$\to i-{\text{C}}_4{\text{H}}_8+{\text{CH}}_4$$ (3)

$${\text{C}}_5{\text{H}}_{12}^{+}\to t-{\text{C}}_4{\text{H}}_9^{+}+{\text{CH}}_3$$ (4)

$$\to i-{\text{C}}_4{\text{H}}_8^{+}+{\text{CH}}_4$$ (5)

$$\to {\text{C}}_3{\text{H}}_5^{+}+{\text{CH}}_3+{\text{CH}}_4$$ (6)

$$\to {\text{C}}_2{\text{H}}_3^{+}+{\text{CH}}_3+{\text{C}}_2{\text{H}}_4$$ (7)

$$\to {\text{C}}_2{\text{H}}_3^{+}+{\text{H}}_2+{\text{CH}}_3+{\text{C}}_2{\text{H}}_4$$ (8)

$$\to {\text{C}}_2{\text{H}}_3^{+}+{\text{H}}_2+{\text{CH}}_3+{\text{CH}}_4$$ (9)

Propyl ions are also produced but at a neopentane pressure of 300 torr [3], $\text{M}\left({\text{C}}_3{\text{H}}_7^{+}\right)/{\text{N}}_{+}=0.01$.

Process 4 is the major mode of fragmentation of ${\text{C}}_5{\text{H}}_{12}^{+}$. At a pressure of 150 torr, $\text{M}\left(t-{\text{C}}_4{\text{H}}_9^{+}\right)/{\text{N}}_{+}$ has been estimated [3] to be 0.75, while $\text{M}\left(t-{\text{C}}_4{\text{H}}_9^{+}\right)/{\text{N}}_{+}$ has been estimated at 0.03. Concrete information concerning relative importances of the neutral dissociative processes is not available. Far ultraviolet photolysis experiments indicate however that process 1 is more important than processes 2 or 3. The total number of neutral electronically excited molecules (${\text{C}}_5{\text{H}}_{12}^{*}$) per positive ion, N_ex/N_i, in the gas phase radiolysis is not exactly known. On the basis of low dose rate experiments an estimated value of 0.2 has been assigned to N_ex/N_i (number of neutral excited molecules per ion pair). Later in the discussion an attempt is made to derive more definitive information concerning the role of the neutral dissociative processes.

4.1. Yields of Fragment Ions

Partially labeled hydrocarbon products such as C₂D₅H, C₂D₃H, and C₃D₅H which were also seen in low dose rate experiments carried out in the presence of NO (tables 2 and 3), can be axcribed to the exothermic hydride ion transfer reactions:

$${\text{C}}_2{\text{D}}_5^{+}+{\text{neo-C}}_{\text{5}}{\text{H}}_{\text{12}}\to {\text{C}}_2{\text{D}}_5{\text{H + C}}_5{\text{H}}_{11}^{+}$$ (10)

$${\text{C}}_2{\text{D}}_3^{+}+{\text{neo-C}}_{\text{5}}{\text{H}}_{\text{12}}\to {\text{C}}_2{\text{D}}_3{\text{H + C}}_5{\text{H}}_{11}^{+}$$ (11)

$${\text{C}}_3{\text{D}}_5^{+}+{\text{neo-C}}_{\text{5}}{\text{H}}_{\text{12}}\to {\text{C}}_3{\text{D}}_5{\text{H + C}}_5{\text{H}}_{11}^{+}$$ (12)

Because the rate constants of these reactions are approximately 10⁻⁹ cm³/molec · s[13], more than 99 percent of the ${\text{C}}_2{\text{D}}_5^{+}$, ${\text{C}}_2{\text{D}}_3^{+}$ and ${\text{C}}_3{\text{D}}_5^{+}$ will react with neopentane even at the highest dose rates used in this study.

If, as in the earlier low dose rate studies [3], one accepts that in fast ion-molecule reactions, isotope effects can be ignored, maximum ion pair yields of the three reactant ions can be obtained by multiplying the sums of ion pair yields of C₂D₅H + C₂DH₅, C₂D₃H + C₂DH₃, and C₃D₅H + C₃DH₅ by a factor of two. The values thus obtained are given in the last column of tables 2 and 3. They compare rather well with those derived from low dose rate experiments carried out at a lower pressure of neopentane in the presence of NO as a free radical interceptor. At low dose rates, low concentrations of NO or O₂ will effectively remove free radicals from the system, and thus prevent contributions of radical-radical interactions to the partially labeled hydrocarbon products. Nitric oxide or oxygen are obviously less efficient as free radical scavengers under our pulse radiolysis conditions. At high dose rates radical-radical interactions will compete with radical-scavenger molecule ractions. Nevertheless, using a 3 ns pulse and a dose of 1.8 eV/g per pulse, the methyl radicals which, according to processes 1 through 8, are the most abundant alkyl radicals produced during the pulse, can be removed to a significant extent upon addition of 5 mol percent NO to neopentane. This is demonstrated by the pronounced drop of M(CH₄)/N₊ and M(C₂H₆)/N₊ as a function of the pressure of NO as seen in figure 1.

Figure 1. The effect of NO on the major hydrocarbon products formed in the pulse radiolysis of neopentane (3ns, 3.6 eV/g × 10¹⁹).

Pressure of neopentane: 162 torr.

Furthermore the CH₃CD₃ which can be unambiguously ascribed to the radical combination reaction

$${\text{CH}}_{\text{3}}+{\text{CD}}_{\text{3}}\to {\text{CH}}_{\text{3}}{\text{CD}}_{\text{3}}$$ (13)

diminishes sharply with respect to the other partially deuterium labeled ethanes when NO is added (table 2). The ion pair yield of 0.047 for the ethyl ions derived from this experiment can be considered as more reliable than the 0.060 calculated from the isotopic ethane distributions observed in the unscavenged experiments (table 2). The value obtained for M(ethyl ion)/N₊ is bracketed by the low dose rate values of 0.068 and 0.035 obtained at neopentane pressures of 30 and 300 torr respectively [3]. The drop of $\text{M}\left({\text{C}}_2{\text{H}}_5^{+}\right)/{\text{N}}_{+}$ with pressure has been ascribed [3] to collisional quenching of process 7.

4.2. Yields of Molecular CH₄ and CH₃

The total yield of molecular methane formed by the fragmentation processes 3, 5, and 6 can be estimated from the isotopic analysis given in table 2 assuming one molecule of CD₄ or CH₄ is being produced for each molecule of CH₃D or CD₃H. In the pulse radiolysis of pure neo-C₅D₁₂-neo-C₅H₁₂ mixtures there is considerable isotopic mixing (table 2) due to the combination reactions

$${\text{CH}}_{\text{3}}+\text{D}\to {\text{CH}}_{\text{3}}\text{D}$$ (14)

$${\text{CD}}_{\text{3}}+\text{H}\to {\text{CD}}_{\text{3}}\text{H}$$ (15)

Addition of 5 mol percent NO drastically reduces the total methane yield (fig. 1), and more specifically, the relatively abundances of CD₃H and CH₃D in the methane fraction. From such an experiment, a reliable value of 0.16 is obtained for the ion pair yield of molecular methane. Because of the large relative yields of CH₃D and CD₃H the value for molecular methane derived from the isotopic distribution observed in pure neo-C₅H₁₂-neo-C₅D₁₂ mixtures is obviously less reliable. In the C₅D₁₂−C₅H₁₂−SF₆ experiment, where the total methane yield (table 1) and the relative contributions of CD₃H and CH₃D, are lower, good agreement is obtained. Taking $\text{M}(i-{\text{C}}_4{\text{H}}_8^{+})/{\text{N}}_{+}=~0.03$ [3] (process 5) and $\text{M}\left({\text{C}}_3{\text{H}}_5^{+}\right)/{\text{N}}_{+}=0.044$ (process 6) more than half of the yield of molecular methane remains unaccounted for. If one tentatively accepts that as indicated by the 70 eV mass spectral pattern ${\text{C}}_3{\text{H}}_3^{+}$ is slightly less abundant than ${\text{C}}_2{\text{H}}_3^{+}$, one obtains a value approximately 0.10 for the molecular methane produced by fragmentation of ${\text{C}}_5{\text{H}}_{12}^{+}$.

M(CH₃)/N₊ can be estimated by summation of all the products which can be ascribed to combination of CH₃ with other radicals in the system

$${\text{CH}}_{\text{3}}+\text{R}\to {\text{CH}}_{\text{3}}\text{R}$$ (16)

where R = H, CH₃, (CH₃)₃CCH₂. In view of the absence of C₃H₃D₃ in the propylene fraction a reaction such as

$${\text{C}}_2{\text{H}}_3{\text{(C}}_{\text{2}}{\text{D}}_3{\text{)+CD}}_{\text{3}}{\text{(CH}}_{\text{3}}\text{)}\to {\text{C}}_3{\text{H}}_3{\text{D}}_3$$ (17)

does not occur and vinyl radicals are therefore not produced in the pulse radiolysis of pure neopentane. Similarly, the negligible contribution of C₄D₅H₃ to the isobutene fraction and the absence of 1-butene as a product indicates that allyl radicals play a very minor role. Finally, the minor yield of propane (see Results) points to a low abundance of ethyl radicals which if formed would mainly combine with CH₃.

After taking into account the contributions of molecular methane and ${\text{C}}_2{\text{H}}_5^{+}$ to the methane and ethane fractions respectively, a value of 1.48 ± 0.02 is obtained for M(CH₃)/N₊ in the pulse radiolysis of pure neopentane. A slightly higher value (1.5 ± 0.02) is obtained if one assumes that C₃H₈ and i-C₄H₁₀ are produced entirely by combination of CH₃ radicals with C₂H₅ and C₃H₇ respectively. A change in pulse duration, dose or dose rate has, within experimental error, no effect on the estimated yield of CH₃ radicals. The value for M(CH₃)/N₊ obtained here is in close agreement with the value of 1.45 reported for M(CD₃H)/N₊ produced in the low dose rate radiolysis of neo-C₅D₁₂ carried out in the presence of H₂S as a free radical interceptor [3]. Accepting [3] a value of 0.03 for $\text{M}\left(\text{i}-{\text{C}}_4{\text{H}}_8^{+}\right)/{\text{N}}_{+}$ the total ion pair of CH₃ resulting from the decomposition of the parent ion (reactions 4, 6 through 9) cannot be higher than 0.97. Since the total yield of methyl radicals derived above is ~ 1.5, it follows that CH₃ radicals are produced from other sources, such as the decomposition of electronically excited neopentane (processes 1 and 2) and of internally excited radicals produced by neutralization.

In order to verify the modes of formation of the methane and ethane products, the pulse radiolysis of neo-C₅H₁₂ containing from 1 to 6.1 mol percent CD₃I (table 4) was investigated. Addition of CD₃I has the effect of introducing CD₃ radicals to the system, mainly by the dissociative electron capture step

$${\text{CD}}_3\text{I}+e\to {\text{CD}}_3+{\text{I}}^{-}$$ (18)

Provided the concentration of CD₃I is sufficiently low, there should be no interference with hydride ion transfer reaction,

$${\text{C}}_2{\text{H}}_5^{+}+\text{neo}-{\text{C}}_5{\text{H}}_{12}\to {\text{C}}_2{\text{H}}_6+{\text{C}}_5{\text{H}}_{11}^{+}$$ (19)

and the free radicals should react in the same way as in pure neopentane. As expected on the basis of the free radical combination reaction discussed above, the methane is seen to consist exclusively of CD₃H and CH₄, while the ethane fraction contains, C₂D₆, CD₃CH₃, C₂H₆ and no other partially deuterated ethanes. If one now subtracts from M(CH₄)/N₊ the contribution of molecular methane, and from C₂H₆ the contribution due to the reaction 19, one obtains consistent values for CH₃/CD₃ as calculated from the product ratios given in the last three columns of table 4. It follows that the mechanism proposed for the formation of methane and ethane is well substantiated.

4.3. Formation of Hydrogen

Apparently, the majority of the H-atoms react with CH₃ to produce methane. On the basis of the results presented in tables 1 and 2, a value of 0.9 can be ascribed to the ion pair yield of H-atoms reacting in this manner. In addition, one has to consider the H₂ which is observed in all experiments. The mechanism by which this product is formed is not immediately obvious. Although the relatively low yield of HD observed in the pulse radiolysis of C₅H₁₂−C₅D₁₂ mixtures, (see Results) indicates that some of the hydrogen is eliminated unimolecularly from excited intermediates, a considerable fraction must be produced via bimolecular reactions. At least some of the HD may be formed in the hydrogen abstraction reaction

$${\text{H}}^{*}+{\text{C}}_5{\text{H}}_{12}\to {\text{H}}_2+{\text{C}}_5{\text{H}}_{11}$$ (19)

where H* designates a translationally excited atom. The presence of 2,2-dimethylbutane (M/N₊ = 0.055 ± 0.05) can be explained by the combination of the resulting neopentyl radicals with methyl radicals. (The yield of this product is reduced in the presence of 5 percent NO, confirming the participation of a radical mechanism in its formation.)

Combination of H-atoms is excluded as a major source of HD, in view of the fact that addition of electron scavengers such as SF₆ and CD₃I reduces M(H)/N₊, (as indicated by the sharp drop of M(CH₄)/N₊ in these experiments) (table 1), but has no noticeable effect on either the yield of hydrogen or its isotopic composition.

4.4. Neutralization

The ions which under the present pulse radiolysis conditions will undergo homogeneous neutralization are: ${\text{C}}_4{\text{H}}_9^{+}$, ${\text{C}}_4{\text{H}}_8^{+}$, ${\text{C}}_3{\text{H}}_3^{+}$ produced in the unimolecular fragmentation steps, and ${\text{C}}_5{\text{H}}_{11}^{+}$ produced via the hydride ion transfer reactions involving ${\text{C}}_2{\text{H}}_3^{+}$, ${\text{C}}_2{\text{H}}_5^{+}$, and ${\text{C}}_3{\text{H}}_5^{+}$. The total ion pair yield of ${\text{C}}_5{\text{H}}_{11}^{+}$ can be estimated at 0.12 ± 0.01 and that of $t-{\text{C}}_4{\text{H}}_9^{+}$ at 0.80. The remaining 0.08 being accounted for by $t-{\text{C}}_4{\text{H}}_8^{+}\left(~0.03\right)$, ${\text{C}}_3{\text{H}}_7^{+}(~0.01)$, and ${\text{C}}_3{\text{H}}_3^{+}(~0.03)$. It is clear from the quantitative results presented in table 1 that the mechanism of neutralization of the $t-{\text{C}}_4{\text{H}}_9^{+}$ ion can be represented in the same way as recently proposed for the liquid phase radiolysis of neopentane [14]:

$$t-{\text{C}}_4{\text{H}}_9^{+}+e\to t\cdot {\text{C}}_4{\text{H}}_9^{*}\mathrm{\Delta }H=-5.17\text{eV}$$ (20)

$$t-{\text{C}}_4{\text{H}}_9^{*}\to i-{\text{C}}_4{\text{H}}_8+\text{H}$$ (21)

$$\to {\text{C}}_3{\text{H}}_6+{\text{CH}}_3$$ (22)

The $t-{\text{C}}_4{\text{H}}_9^{*}$ radical produced initially will possess 5.17 eV in internal energy [15] and will therefore dissociate essentially instantaneously [16]. Step 21, which on the basis of pyrolysis studies [17] requires an activation energy of 1.9 eV, is the expected mode of decomposition of internally excited t-C₄H₉ radicals. Decomposition step 22 which, in pyrolysis carried out at temperatures up to 857 K, plays a very minor role, is however of importance in the low intensity photolysis of neopentane at 147 and 123.6 nm [3] as well as in the flash photolysis (table 1). The internal energy content of the intermediate $t-{\text{C}}_4{\text{H}}_9^{*}$ radical, which in the photolysis is produced by cleavage of a C–C bond in neopentane, is comparable to that produced in the neutralization step. It is matter of conjecture if the $t-{\text{C}}_4{\text{H}}_9^{*}$ isomerizes first to the $i-{\text{C}}_4{\text{H}}_9^{*}$ structure or if a CH₃CCH₃ diradical is produced by a simple C–C cleavage of the $t-{\text{C}}_4{\text{H}}_9^{*}$ intermediate followed by isomerization to propylene. It is of interest, however, that isobutyl radicals have been reported in the liquid phase radiolysis [18] where some of the ${\text{C}}_4{\text{H}}_9^{*}$ may be collisionally stabilized. The branching ratio of steps 21 and 22 can be estimated by the decrement of M(C₃H₆)/N₊ noted when an electron scavenger such as SF₆, CCl₄, or CD₃I is added to neopentane (see table 1). In view of the high electron capture cross section of these molecules [19], a few mole percent should be sufficient to capture the majority of the electrons under the pulse radiolysis conditions. This is indicated by the observation (table 4) that in the case of the least efficient electron scavenger, CD₃I, M(CD₃)/N₊ which is mainly produced by the dissociative electron attachment process 18 reaches values of 0.81 and 1.05 upon addition of 1 and 3 mol percent CD₃I respectively. Neutralization of a positive ion by collision with a negative ion rather than an electron will impart less energy to the neutral entity formed in the process. The electron affinity of the scavenger molecule will result in the formation of a negative ion with a greater heat of formation than the electron, so the neutralization reaction is less exothermic, and the energy content of the neutralized species will be lowered. As is indicated by the reduction of M(H)/N₊ in the presence of an electron scavenger, the energy may be further reduced because of the occurrence of chemical reactions such as

$${\text{C}}_4{\text{H}}_9^{+}+{\text{I}}^{-}\to \text{HI}+i-{\text{C}}_4{\text{H}}_8$$ (23)

$${\text{C}}_4{\text{H}}_9^{+}+{\text{SF}}_6^{-}\to \text{HF}+{\text{SF}}_5+i-{\text{C}}_4{\text{H}}_8$$ (24)

Reduction of hydrogen atom production by electron scavengers have also been reported in numerous low dose rate radiolysis studies [20] of alkanes and cycloalkanes.

Addition of CD₃I reduces M(C₃H₆)/N₊ from 0.21 ± 0.05 to 0.10, independent of the concentration of CD₃I. CCl₄ has the same effect as CD₃I, while SF₆ seems to reduce the yield somewhat less. Accepting, on the basis of the reduction in the yield of C₃H₆ in the presence of CD₃I, a value of 0.11 for the ion pair yield of reaction 22, one can assume that the remaining $t-{\text{C}}_4{\text{H}}_9^{+}$ ions (M/N₊ = 0.69) lead to the production of i-C₄H₈ (reaction 21).

The $i-{\text{C}}_4{\text{H}}_8^{*}$ produced in the neutralization of the $i-{\text{C}}_4{\text{H}}_8^{+}$ ion

$$i-{\text{C}}_4{\text{H}}_8^{+}+e\to i-{\text{C}}_4{\text{H}}_8^{*}$$ (25)

will possess 9.23 eV in energy above the ground state of the i-C₄H₈ molecule. It may be assumed that this molecule will dissociate in a manner similar to that observed for $i-{\text{C}}_4{\text{H}}_8^{*}$ molecules excited by absorption of 10 eV photons, namely:

$$i-{\text{C}}_4{\text{H}}_8^{*}\to {\text{CH}}_2=\text{C}={\text{CH}}_2+\text{H}+{\text{CH}}_3$$ (26)

$$\to {\text{CH}}_3\text{C}\equiv \text{CH}+\text{H}+{\text{CH}}_3$$ (27)

Both decompositions occur via a ${\text{C}}_4{\text{H}}_7^{*}$ and/or ${\text{C}}_3{\text{H}}_5^{*}$ intermediate. Methylacetylene and allene are produced in the pulse radiolysis of neopentane (table 1), and it can be seen that addition of an electron scavenger reduced M(C₃H₄)/N₊ from 0.075±0.005 to 0.040±0.005. The decrement corresponds to the ion pair yield of $i-{\text{C}}_4{\text{H}}_8^{+}$. The residual C₃H₄ could be produced by decomposition of highly excited neutral fragments produced in the decomposition of electronically excited neopentane. Both allene and methyl-acetylene are minor products in the flash photolysis of neopentane (table 1).

The mechanism of neutralization of the ${\text{C}}_5{\text{H}}_{11}^{+}$ ion is rather straightforward. 2-Methyl-l-butene and 2-methyl-2-butene are products in the pulse radiolysis of neopentane (see Results). The fact that their yields are not reduced by addition of NO shows that they are not formed by a radical combination reaction. Also none of these products was observed in the flash photolysis or low intensity photolysis of neopentane. A logical interpretation involves the neutralization of ${\text{C}}_5{\text{H}}_{11}^{+}$ ions with the structure ${\left({\text{CH}}_3\right)}_2{\text{CCH}}_2{\text{CH}}_3^{+}$

$${\left({\text{CH}}_3\right)}_2{\text{CCH}}_2{\text{CH}}_3^{+}+e\to {\left({\text{CH}}_3\right)}_2\text{C}={\text{CHCH}}_3+\text{H}$$ (28)

$$\to {\text{CH}}_3{\text{CH}}_2\text{C}\left({\text{CH}}_3\right)={\text{CH}}_2+\text{H}$$ (29)

process 29 occurring with a probability which is a factor of 2.5 times higher than 28. In view of the fact that in the pulse radiolysis of pure neopentane M(C₅H₁₀)/N₊ = 0.09 it can be assumed that essentially all ${\text{C}}_5{\text{H}}_{11}^{+}$ ions produced in the hydride ion transfer reactions acquire the stable $t{\text{-C}}_5{\text{H}}_{11}^{+}$ structure prior to neutralization. Addition of SF₆ raises M(C₅H₁₀)/N₊ to 0.12 ± 0.01 in good agreement with the ion pair yield of the ${\text{C}}_5{\text{H}}_{11}^{+}$ ions (0.12). It can be surmised that in the absence of SF₆ a fraction (M/N₊ ~ 0.03) of the ${\text{C}}_5{\text{H}}_{11}^{+}$ are neutralized in the following manner:

$${\left({\text{CH}}_3\right)}_2{\text{CCH}}_2{\text{CH}}_3^{+}+e\to {\left({\text{CH}}_3\right)}_2{\text{CCH}}_2+{\text{CH}}_3$$ (30)

The mechanism of neutralization involving ${\text{SF}}_{\overline{6}}$ is probably similar to reaction 24.

Neutralization of the remaining minor ions, ${\text{C}}_3{\text{H}}_3^{+}$ and sec-${\text{C}}_3{\text{H}}_7^{+}$ can not be elucidated. They will probably contribute to formation of highly unsaturated products. The value of M(C₂H₂)/N₊, for instance, is reduced from 0.051 to ~ 0.035 upon addition of electron scavengers.

4.5. Estimation of the Yield of Electronically Excited Neopentane Molecules

Table 5 represents a summary of the ion pair yields of free radicals and molecular products, which have been accounted for by the unimolecular dissociation of the parent ion and the major neutralization reactions proposed in the previous section. The last row of the table shows the yields of the fragments which are not yet accounted for. An ion pair yield of 0.06 must be ascribed to molecular ethylene which does not enter into any of the reactions listed so far. Even though not all of the ionic processes have been exhaustively covered, the yields given in the last row provide a reasonable balance between the C₁ products on one hand and the C₂, C₃, and C₄ products on the other hand, if one takes into account that two C₁ fragments will be formed for each C₃ fragment. If one tentatively ascribes the unaccounted for products to decomposition of electronically excited neopentane molecules produced by electron impact a value of 0.28 is obtained for N_ex/N_i. This estimated value may admittedly require some revision wherever the minor unexplained reaction steps can be accounted for. It falls however in the same range as the values derived from conventional low dose rate radiolysis experiments on neopentane [3] as well as other alkanes [21].

Table 5.

Methyl radicals and hydrocarbon molecules produced in unimolecular dissociation and neutralization reactions

	CH3	CH4	C2H2	M/N+		C3H6	i-C4H8	C5H10
				C2H4	C3H4
neo-C5H12 + → C4H9+ + e → C5H11+ + e → C4H8+ + e →	0.97	0.10
	.11					0.11	0.69
	.03						0.03	0.09
	.03				0.03
Total:	1.14	0.10			0.03	0.11	0.72	0.09
Unaccounted:	0.34	0.06	0.035	0.06	0.04	0.06	0.13